In situ formation of low molecular weight organogelators for slick solidification

We have investigated the in situ formation of Low Molecular Weight Organogelator (LMWO) molecules in oil-on-water slicks through dual reactive precursor injection. This method alleviates the need for any carrier solvent or prior heating, therefore reducing the environmental impact of LMWOs, giving instantaneous gelation, even at low temperatures (−5 °C). We show minimal leaching from our gels into the water layer.


Table of Contents
: Inversion tests for 1-octadecene slicks on deionised water of: left: 1 and right: 2. Note that at 5 wt.% gelator concentration, with a larger volume of 1-octadecene (not seen with conditions described in Figure S4) , the integrity of 2 gel was compromised in the inversion test. Figure S2: Photographs showing Compound 1 gelling a 1-octadecene slick at room temperature (a) and b)) and at -5 °C (c)). 1-octadecene slicks were placed above seawater and the two flasks are shown for comparison in d).    Table S1: Gelation tests were carried out at 4 wt% w/v LMWO on 4 g of selected oils placed on 4 ml seawater, forming a thick slick. The sample was then placed in an ice-bath and cooled to -5 ºC before injection of LMWO precursors. The gelation properties of the 2 compounds in oils of varying polarity. G indicates a gel formed (survived the inversion test) and the CGC estimated from inversion testing is given in brackets (units% w/v). NG indicates that no gel was formed, G* indicates that a gel was formed under the above conditions but did not survive the inversion test (with an estimate of CGC given in brackets) and D indicated that the LMWO precursors dissolved and did not gel.    Table S1, with toluene, hexan-1-ol, kerosene, Mobil Super 3000 motor oil, Vegetable oil and the octadec-1-ene control on seawater (Brighton, UK).   Table S1.

Part I: Preparation of 1 in situ and analysis of resulting gel
N'-(4-Methylphenyl)-N,N-dipropan-2-ylurea, compound 1, made in situ, by co-adding from opposite ends of the vessel neat 1-isocyanato-4-methylbenzene (p-tolyl isocyanate) and neat N-(propan-2-yl)propan-2-amine (diisopropylamine) to the oil layer floating on top of a water layer. Gelation off the whole oil layer occurred within 30 seconds of the addition of the chemical precursors.
The NMR spectroscopy:

Sample preparation
Samples (ca. 30 dry gelator or 100mg gelator + oil) were prepared in CDCl 3 (Goss Scientific Instruments Ltd., Crewe, UK) and spiked with a trace amount of tetramethylsilane (Merck, Gillingham, UK) as the 0 ppm internal reference for 1 H spectra.

NMR analysis
1D ( 1 H, 13 C DEPT-Q) and 2D ( 1 H-1 H COSY, 1 H-13 C HSQC) NMR experiments were carried out to confirm the presence of gelator 1. The parameter sets used to acquire the datasets were the standard ones provided by the spectrometer manufacturer. The data were processed using TopSpin 3.6.1 (Bruker BioSpin GmbH, Rheinstetten, Germany).
1D 1 H (base optimised): 16 scans, 64K complex data points, 8.2KHz spectral width. 1D proton data were fast Fourier transformed after applying a single exponential apodisation window (lb = 0.2Hz) and zero filling to 64K real data points. The 1D 1 H spectrum was automatically referenced to internal TMS, automatically phase-corrected with manual adjustment of the phase as required, and automatically baseline corrected (using polynomial degree 5 curve fitting). Subsequent 1D and 2D spectra were referenced using the corresponding TMS-referenced 1 H spectrum as per IUPAC recommendations. [1] 1D 13 C DEPT-Q (base optimised): 256 scans, 64K complex data points, 24KHz spectral width. 1D carbon data were fast Fourier transformed after applying a single exponential apodisation window (lb = 1.0Hz) and zero filling to 64K real data points. The spectrum was automatically phase-corrected, with manual adjustment of the phase as required, and automatically baseline corrected (using polynomial degree 5 curve fitting).

Figure 5: 1 H NMR spectra (CDCl 3 ) of gelator-1 made in-situ (2 attempts) vs neat
Green = gelator 1 made by solvent-free reaction of diisopropylamine with p-tolyl isocyanate. Red = gelator 1 made by addition of diisopropylamine to p-tolyl isocyanate mixed with oil layer floating on water layer. Blue = gelator 1 made by coaddition of diisopropylamine and p-tolyl isocyanate into oil layer floating on water layer.

Part II: Preparation and analysis of 2
N-dodecyl-N'-(4-methylphenyl)-urea, compound 2, made in situ, by adding neat 1-isocyanato-4-methylbenzene (p-tolyl isocyanate) to the n-hexane on top of a water layer followed by neat dodecan-1-amine (dodecylamine). Gelation of the whole organic layer occurred within 30 seconds of the addition of the chemical precursors.

NMR analysis
1D ( 1 H, 13 C DEPT-Q) and 2D ( 1 H-1 H COSY, 1 H-13 C HSQC) NMR experiments were carried out to characterise gelator 2. The parameter sets used to acquire the datasets were the standard ones provided by the spectrometer manufacturer. The data were processed using TopSpin 3.6.1 (Bruker BioSpin GmbH, Rheinstetten, Germany).
1D 1 H (base optimised): 16 scans, 64K complex data points, 12KHz spectral width. 1D proton data were fast Fourier transformed after applying a single exponential apodisation window (lb = 0.2Hz) and zero filling to 64K real data points. The 1D 1 H spectrum was automatically phase-corrected with manual adjustment of the phase as required, and automatically baseline corrected (using polynomial degree 5 curve fitting). The 1 H spectrum was referenced to the solvent residual signal at 2.50ppm [2] . Subsequent spectra were indirectly referenced to the 1 H spectrum.
1D 13 C DEPT-Q (deptqgpsp.2, base-optimised): 256 scans, 64K complex data points, 36KHz spectral width. 1D carbon data were fast Fourier transformed after applying a single exponential apodisation window (lb = 1.0Hz) and zero filling to 64K real data points. The spectrum was automatically phase-corrected, with manual adjustment of the phase as required, and automatically baseline corrected (using polynomial degree 5 curve fitting).
degree 5 curve fitting). Finally the spectrum was displayed in magnitude mode (F2 magnitude calculation) to improve signal intensity.

Sample preparation
Sea/river water (10 ml) was placed in a vial and 1-octadecene (1 ml) was layered on top of it. The precursors to the gelator were co-added to the organic phase to in order produce a gel with a 10 wt% gelator:oil ratio. After approximately 15 minutes, the aqueous layer below the newly formed gel was gently homogenised and sampled for NMR analysis. Each NMR tube was assembled using the aqueous layer from the gelation experiment (0.540 ml), D 2 O (0.060 ml) and a bolus (0.030 ml) of dimethyl sulfone (Merck, Gillingham, UK) stock solution (9.10 mg in 1.000 ml) as internal standard. Prior to NMR analysis, samples were spiked with a trace amount of 3-(Trimethylsilyl)propionic acid-d4 sodium salt (TSP) (Merck, Gillingham, UK) as the internal reference (0 ppm).

NMR analysis
1D 1 H spectra were acquired with water suppression using the manufacturer supplied pulse program noesygppr1d, which uses presaturation during relaxation and mixing time. The water peak position and optimal pulse length were automatically adjusted for each sample which was maintained at a constant 298º K during the measurements. Relevant acquisition parameters included a receiver gain (RG) set to 128, 64 scans, 64K complex data points and 12.3 KHz spectral width. The acquisition time was 2.65 seconds and the relaxation delay (D1) was 30 seconds to allow relaxation of the protons between pulses. The quantitation standard dimethyl sulfone was chosen as it is fully soluble in water and its well resolved singlet did not overlap with the peaks of interest. These conditions were selected to achieve quantitative integrals and yield appropriate signal to noise ratios for the analytes of interest. [3,4] The data were processed using TopSpin 3.6.1 (Bruker BioSpin GmbH, Rheinstetten, Germany) applying fast Fourier transform to the data after applying a single exponential apodisation window (lb 1.0Hz, to minimise noise) and zero filling to 64K real data points. The spectra were automatically referenced to internal TSP, automatically phase-corrected with manual adjustment of the phase as required. The baseline was corrected automatically using polynomial degree 5 curve fitting, prior to manual integration of the peaks of interest. In order to unequivocally match aromatic signals to the individual components detected in the mixture, water-suppressed 2D COSY spectra were acquired using the manufacturer's supplied pulse program "cosygpprqf". Relevant parameters were 16 scans, 4096 x 256 complex data points and a spectral width of 8.4KHz (optimised from 1 H). The centre of the spectrum was obtained from the 1 H spectrum and used as the water suppression frequency. Processing involved sine window functions F2 (lb = 1.0Hz, GB = 0, SSB = 0) and in F1 (lb = 0.3Hz, GB = 0.1, SSB = 0) and zero filling in F1 to produce a 2048 * 2048 real data points.
The 1 H NMR spectra obtained are presented staked together on Figure 23. NMR signals that were looked for in the leaching experiments are given in Table 1. The quantitation results are given in Table 2.   The values reported above are those determined from a single NMR sample on one peak for each analyte where a signal to noise ratio above 240 was measured (unless otherwise noted). When other peaks were available for integration which had S/N in as low as 77 the resulting concentrations were essentially the same as those presented above. The results for 2 on river water were determined on an aged sample. Aging seemed to have little effect on quantitation results for 2 (data in spreadsheet provided with this supporting information).